Serotonin has been implicated in a large number of processes including the regulation of sleep, appetite, mood, aggression, perception, memory, and anxiety. At least 14 separate 5-HT receptors have evolved, which are divided into seven main families. Not surprisingly, alterations of 5-HT receptor activity have been shown to occur in many psychiatric diseases including anxiety, depression, eating disorders, schizophrenia, personality disorders, and many drug-induced psychotic states. Additionally, a number of effective psychopharmacologic agents for diseases as diverse as schizophrenia and anxiety have been developed which either specifically alter brain levels of serotonin or bind to 5-HT receptor subtypes. We propose to use a novel tripartite approach to develop an understanding of the relationships between ligand structure, neurotransmitter receptor structure, and ligand-receptor association. Specifically, we intend to elucidate the molecular determinants of the interactions between 5-HT2 receptors and a series of novel tricyclic antagonists using an integrated approach that combines information from site-directed mutagenesis and ligand structure-activity relationships (SAR) to refine hypothetical three-dimensional (3D) receptor models. The first and most basic event that determines the pharmacological activity of an agent is the association of a ligand with the receptor. The ultimate pharmacological outcome is a result of receptor activation or deactivation following formation of the ligand-receptor complex. Since there are no direct experimental structures for the membrane bound G-protein coupled receptors, the molecular details of ligand-receptor structure can only be investigated indirectly by examining ligand SAR and receptor SAR by site-directed mutagenesis. Computational chemistry and molecular modeling provide means to evaluate and organize indirect data into a hypothetical 3D framework at the atomic level of detail. Such 3D models provide a means to not only organize experimental observations but also to generate testable hypotheses concerning ligand-receptor interactions. We will synthesize and evaluate compounds designed specifically on the basis of receptor models to test the importance of certain amino acid residues for ligand binding and receptor function, thus testing model accuracy. The affinities and functional properties of the designed target compounds will be evaluated with both the native and selected mutant receptors. This is one of the first times that a combined approach, utilizing receptor modeling, model-specific ligand design, and model-directed mutagenesis, has been applied to 5-HT receptors.